Introduction
When whole milk enters the calf’s abomasum, casein coagulates under the combined effects of acid, calcium, and the enzyme chymosin. Much of the casein and fat are retained in a curd while the liquid whey phase leaves the abomasum more rapidly. This is one of the most distinctive features of milk digestion in the young calf.
For many years, that observation led to a simple conclusion: a good calf milk replacer should clot. Milk replacers that did not clot were often considered inferior, and the inclusion of dried skim milk became a convenient way to provide casein and encourage curd formation. The difficulty is that the comparison of a skim-based formula with a whey-based formula changes much more than clotting.
If we want to know whether the curd itself improves digestion, health, or growth, the cleanest experiment is not “skim versus whey.” It is to feed essentially the same skim-milk formula and prevent one treatment from coagulating. Several older studies did exactly that, and their results remain remarkably relevant.
Why skim versus whey is a complicated comparison
Consider the recent Górka and Budziński (2026) study reviewed in Calf Note #286. The three milk replacers were formulated to contain similar protein, fat, energy, and essential amino acid concentrations, but increasing skim milk necessarily changed the rest of the ingredient matrix. The 0% skim formula contained substantial WPC and delactosed whey; the 25% formula contained skim milk plus lower WPC; and the 50% formula contained no WPC and instead included sweet whey.
Therefore, if calves respond differently, the biological cause could be the casein:whey ratio, clot formation, processing history of the individual ingredients, mineral composition, peptide profile, gastric emptying, or another characteristic of the complete formula. A performance trial can tell us which formula worked better. It cannot automatically tell us why.
What does a curd actually do?
Curd formation changes the timing of nutrient delivery. Casein and much of the fat can remain in the abomasum while lactose, minerals, water, and soluble whey proteins move into the intestine more quickly. The curd is gradually digested and eroded, producing a more prolonged flow of some nutrients to the small intestine.
That different pattern is real and physiologically important. But a slower pattern of delivery is not automatically a greater amount of nutrient digested. To show a nutritional advantage, clotting should improve total or ileal digestibility, nitrogen retention, health, feed efficiency, or growth. The experimental evidence for those outcomes is much less convincing than the evidence that clotting changes gastric kinetics.
A clever way to isolate clotting
Petit and colleagues developed an oxalate-sodium hydroxide buffer that bound calcium and prevented casein coagulation. The important feature of this approach was that the underlying milk replacer remained based on skim milk. Thus, the researchers could compare a clotting and a nonclotting version of essentially the same protein source rather than replacing casein with whey or plant protein.
In an early validation study, Petit et al. (1987) showed that the buffer itself did not depress dry matter, protein, or fat digestibility in a high-heat skim-milk replacer. They then used low-heat skim milk powder to compare normal clotting with experimentally suppressed clotting.
Petit et al. (1988) found similar digestibility of dry matter, protein, and fat in calves fed clotting or nonclotting milk replacer. Preventing the curd did change several postprandial blood measurements – including triglycerides, essential amino acids, and urea – which is consistent with a change in the rate or pattern of nutrient delivery. Yet the amount of the diet ultimately digested was not reduced.
A subsequent study using ileal as well as fecal collection reached a similar conclusion. Clotting altered some aspects of digesta flow, but it did not change nitrogen, fat, or individual amino acid digestibility in a way that would support the idea that coagulation is essential for efficient protein use.
Cruywagen added the feeding-level question
Cruywagen and Horn-Quass (1991) made this question even more interesting. Sixteen Friesland bull calves received a commercial, coagulable CMR. In one treatment, casein coagulation occurred normally; in the other, coagulation was prevented with an oxalate-sodium hydroxide buffer. The daily feeding allowance was then increased each week from 9 to 12, 15, and finally 18% of birth weight.
Table 1. Selected experiments that separated curd formation from the underlying milk protein source.
| Study | Approach | Main result |
| Petit et al., 1988 | Low-heat skim CMR; clotting vs. chemically suppressed clotting | DM, protein, and fat digestibility were similar; blood metabolite patterns changed. |
| Petit et al., 1989 | Same concept with ileal and fecal collection | No effect of clotting on N, fat, or amino acid digestibility. |
| Cruywagen & Horn-Quass, 1991 | Coagulable CMR; curd suppressed; allowance increased to 18% of birth weight/day | No effect on gain, diarrhea, or DM, OM, and CP digestibility. |
| Miyazaki et al., 2010 | All calves fed the same in vitro clotting CMR; curd visualized by ultrasound | 8 of 29 calves formed no visible curd, yet most appeared healthy and blood measures were not impaired. |
By week 4, the calves were receiving CMR equal to 18% of birth weight per day. Body-weight gain was 10.4 kg for the normally clotting treatment and 10.7 kg for the nonclotting treatment. Feed conversion and digestibility were also similar. Crude protein digestibility was numerically 81.0% in the clotting treatment and 85.6% in the nonclotting treatment.
This study directly addresses an important objection to older clotting work: perhaps curd formation matters only when more liquid is fed. Even as the allowance increased substantially, suppressing the curd did not reduce growth or nutrient digestibility. The study still did not reproduce today’s very largest 4- to 5-L individual meals, but at two feedings per day an 18%-of-birth-weight daily allowance was hardly a severely restricted program.
Even a “clotting” CMR does not always clot in the calf
Miyazaki et al. (2010) added another wrinkle. They used ultrasonography to observe the abomasum of 29 preruminant calves after the calves consumed a CMR that had demonstrated clotting ability in an in vitro rennet test. A visible abomasal curd was absent in 8 of the 29 calves at 2 hours after feeding, and it remained absent through 7 hours in those calves.
The surprising part was not only that some calves failed to form a curd. Most of the non-curd-forming calves appeared healthy, consumed their feed, and were vigorous. Serum triglyceride, blood urea nitrogen, and glucose did not show evidence of a major nutritional failure. Thus, clotting is not simply a property of the powder; it is the result of an interaction between the formula and the individual calf’s abomasal environment.
Clotting changes kinetics more clearly than it changes performance
When these experiments are considered together, a fairly consistent picture emerges. Curd formation changes what happens immediately after a meal. It can alter gastric retention, the timing of fat and protein flow, and the appearance of nutrients in blood. Those effects are physiologically real.
What has been much harder to demonstrate is that the curd increases the total amount of protein or fat digested, prevents diarrhea, or produces faster growth. Longenbach and Heinrichs (1998) reviewed this controversy and concluded that many of the poor results historically associated with nonclotting replacers were more likely related to ingredient quality, manufacturing methods, or the inclusion of poorly digested protein sources than to the absence of a curd itself.
This distinction helps explain why the calf industry could move from skim-milk-based formulas toward whey-based formulas without a general collapse in calf performance. Whey-based formulas changed gastric kinetics, but modern high-quality whey proteins were much more digestible than many of the nonclotting ingredients used in early milk replacers.
Does meal size still matter?
The volume question is not completely closed. Vorndran and Steinhoff-Wagner (2025) allowed calves to consume approximately 10% of birth weight in a single meal twice daily. At that scale, differences in osmolality, nutrient density, and gastric emptying could become more important than they are in a 2-L meal. Their study, however, compared dam’s milk with a compositionally different CMR, so clotting could not be isolated.
Cruywagen and Horn-Quass provide stronger mechanistic evidence because they suppressed clotting without replacing the protein source and increased the feeding level. Their lack of a negative response makes it difficult to argue that curd formation suddenly becomes essential whenever daily milk intake increases. Still, a modern factorial study specifically comparing clotting and nonclotting formulas at very large versus small individual meals would be useful.
A more useful way to think about the curd
I do not think the evidence supports saying that clotting is unimportant. It clearly affects digestion kinetics, and those kinetics could matter under particular conditions – especially in very young calves, with unusual meal sizes, or when the quality of the protein ingredients is marginal. But it is equally difficult to defend the statement that a firm curd is required for normal digestion and growth.
The more defensible conclusion is that clotting is one functional characteristic of a milk replacer, not a universal quality test. A CMR that forms a beautiful curd can still contain poorly processed protein, and a high-quality whey-based CMR may not form a curd at all while supporting excellent performance.
That leads directly to the third Note in this series. If percentage skim milk and clotting ability do not reliably predict performance, how should we evaluate the quality of the milk ingredients themselves? Calf Note #288 will focus on processing, heat damage, rennet coagulation tests, WPNI, and other ways to characterize milk powders.
Summary
Comparing skim- and whey-based CMRs changes multiple characteristics simultaneously, so those trials cannot isolate the biological value of curd formation. Experiments that prevented clotting while keeping the underlying skim-milk formula largely unchanged provide a cleaner test. Across studies by Petit and by Cruywagen and Horn-Quass, suppressing curd formation changed nutrient kinetics but generally did not reduce protein digestibility, health, or growth. Cruywagen also increased the daily liquid allowance to 18% of birth weight without revealing a benefit of clotting. Curd formation is real physiology; the evidence that it is required for good calf performance is much weaker.
References
Cruywagen, C. W., and J. G. Horn-Quass. 1991. Effect of curd suppression of a calf milk replacer fed at increasing levels on nutrient digestibility and body mass-gain. S. Afr. J. Anim. Sci. 21:153-156. Link
Górka, P., and W. Budziński. 2026. Effect of inclusion of skim milk in milk replacer on growth performance of newborn dairy calves. J. Dairy Sci. doi:10.3168/jds.2026-29094. Link
Longenbach, J. I., and A. J. Heinrichs. 1998. A review of the importance and physiological role of curd formation in the abomasum of young calves. Anim. Feed Sci. Technol. 73:85-97. doi:10.1016/S0377-8401(98)00130-8. Link
Miyazaki, T., M. Miyazaki, J. Yasuda, and K. Okada. 2010. No abomasal curd formation in pre-ruminant calves after ingestion of a clotting milk replacer. Vet. J. 183:205-209. doi:10.1016/j.tvjl.2008.09.006. Link
Petit, H. V., M. Ivan, and G. J. Brisson. 1987. An oxalate-sodium hydroxide buffer to study the role of milk replacer coagulation in preruminant calves. J. Dairy Sci. 70:2565-2569. doi:10.3168/jds.S0022-0302(87)80325-9. Link
Petit, H. V., M. Ivan, and G. J. Brisson. 1988. Digestibility and blood parameters in the preruminant calf fed a clotting or a nonclotting milk replacer. J. Anim. Sci. 66:986-991. doi:10.2527/jas1988.664986x. Link
Petit, H. V., M. Ivan, and G. J. Brisson. 1989. Digestibility measured by fecal and ileal collection in preruminant calves fed a clotting or a nonclotting milk replacer. J. Dairy Sci. 72:123-128. doi:10.3168/jds.S0022-0302(89)79087-1. Link
Vorndran, A. M., and J. Steinhoff-Wagner. 2025. Effects of high-volume voluntary feed consumption on growth and health of Brown Swiss calves: Continued dam’s milk feeding versus early introduction of milk replacer during the first week of life. J. Dairy Sci. 108:5954-5967. doi:10.3168/jds.2025-26405. Link