Subclinical mastitis (SCM) is a major health problem of dairy animals in India and across the globe. An identification of potential risk factors of SCM can help for efficient udder health management in dairy animals. In this study, apparently healthy cows (HF crossbred: n = 45; Deoni: n = 43) were screened for SCM during different seasons through milk somatic cell count (SCC: reference test using 200 × 103 cells/ml as cut off value), California mastitis test (CMT) and differential electrical conductivity (DEC) test at an organized research farm. SCM positive milk samples (n = 34) were inoculated in selective media for Coliform sp., Streptococcus sp. and Staphylococcus sp. and DNA was isolated (n = 10) for species confirmation by 16s rRNA method. Both bivariate and multivariate models were used for risk assessment. We found the cumulative prevalence of 31 and 65% SCM in Deoni and crossbred cows, respectively. Screening of 328 crossbred cows under field conditions revealed point prevalence of 55% SCM. Multivariate analysis revealed stage of lactation (SOL), milk yield in previous lactation and test day milk yield in Deoni cows, as well as parity and mastitis treatment history in current lactation in HF crossbred cows as risk factors. SOL was a significant factor under field conditions. Receiver operated characteristic curve analysis revealed better accuracy of CMT than DEC. We found more mixed infections due to Staphylococcus sp. and Streptococcus sp. in culture, while 16s rRNA based molecular method revealed lesser-known pathogens associated with SCM. It is concluded that SCM prevalence rate is higher in crossbred than indigenous cows and these breeds have different risk factors for SCM. HF crossbred cows had similar SCM prevalence rate under different farming conditions, where CMT can be used for SCM diagnosis with excellent accuracy. The 16s rRNA method is useful for specific identification of lesser known and emerging mastitis pathogens.