OBJECTIVE To evaluate a 3% solution of household detergent viz., Surf Excel (Surf field mastitis test, SFMT) vis-à-vis California mastitis test (CMT), Whiteside test (WST), somatic cell counts (SCC; cut off limit = 5 x 10(5) cells per millilitre) and bacteriological cultures for the detection of subclinical mastitis in quarter foremilk samples (n=800) of dairy cows and buffaloes. METHODS Culture and SCC were used as gold standards. All tests were evaluated parallel and serial patterns. The sensitivities of SFMT, SCC, culture, CMT and WST in parallel testing were 72.82, 81.55, 87.38, 75.73 and 54.37%, respectively in cows, while 66.22, 79.73, 82.43, 70.27 and 50.00, respectively in buffaloes. RESULTS SFMT was significantly (p<0.05) more sensitive than WST and comparable to CMT in both species. In serial testing, percent specificity of SFMT (87.12 in cow; 85.16 in buffaloes) was significantly (P<0.05) higher than that of CMT (83.33 in cow; 80.64 in buffaloes). The negative predictive values of SFMT (93.50 in cow; 96.35 in buffaloes) differed non-significantly from that of CMT (94.02 in cow; 96.15 in buffaloes). The kappa index between the tests was moderate to perfect both in parallel (0.54 to >0.80) and serial (0.58 to >0.8) testing. CONCLUSIONS On the basis of closely similar diagnostic efficiency of SFMT to CMT in terms of sensitivity, specificity, predictive values and kappa index together with inexpensive and ready availability of SFMT reagent, it tempting to suggest that SFMT can be use as a cheaper, user-friendly alternative animal-side subclinical mastitis diagnostic test in poor countries.