Reducing sugar (RSA) and phenol–sulfuric acid (PSA) assays are commonly used to analyze water-soluble carbohydrates. However, questions have arisen as to their accuracy for measurements of feedstuffs with diverse carbohydrate profiles. This study evaluated the efficacy of RSA and PSA as they would commonly be applied in feed analysis laboratories in measuring a variety of purified carbohydrates. Carbohydrates analyzed were glucose (Glc), fructose (Fru), galactose (Gal), sucrose (Suc), maltose (Mal), lactose (Lac), raffinose (Raf), and inulin (Inu). Variations on the methods used were PSA using Suc (PSA-Suc) or Glc (PSA-Glc) as standard sugars, and RSA with a 50:50 Glc:Fru blend as the standard with four hydrolysis methods: acid hydrolysis with 0.037 M sulfuric acid (RSA-H2SO4) or 0.5M hydrochloric acid (RSA-HCl), or enzymatic hydrolysis with invertase (RSA-Inv) or an enzyme blend including sucrase, α-glucosidase, and β-galactosidase (RSA-EnzBl). Recovery of carbohydrate was calculated on a dry matter (DM) basis as (carbohydrate detected g/kg DM)/(carbohydrate present kg/kg DM), with ‘close to’ complete recovery defined as values falling within the range of 920–1080 g/kg. Monosaccharide recovery did not differ between unhydrolyzed vs. hydrolyzed samples in RSA indicating no destruction of carbohydrate by hydrolysis method. For RSA, recoveries of Glc, Fru, and Gal were 979, 1042, and 706 g/kg, respectively. Such response differences among monosaccharides are inherent to RSA, and can affect carbohydrate recovery values. Methods that provided close to complete recovery by carbohydrate were: PSA-Suc and all RSA for Suc; PSA-Glc and RSA-EnzBl for Mal and Lac; PSA-Suc, RSA-H2SO4, RSA-HCl, and RSA-Inv for Raf; and RSA-H2SO4 and RS-HCl for Inu. None of the assays gave complete recovery of the diverse set of purified carbohydrates. Allowing a range of 920–1080 g/kg for recoveries on individual carbohydrates, RSA-H2SO4 and RSA-HCl would give the closest to complete recovery values for feeds such as forage and soybean in which Suc, Raf, and Inu were important, whereas RSA-EnzBl would be useful in feeds such as forages or dairy products when Suc, Mal, and Lac are of interest. The allowed 920–1080 g/kg range of acceptable recoveries addresses the point that given very diverse carbohydrate complements of feeds, these assays will not be extremely precise, but may still be serviceable for diet formulation. The most accurate measurements will be achieved by selection of detection method, hydrolysis method, and carbohydrate standard to give greatest recovery of predominant carbohydrates in feedstuffs.