Several osteoporosis risk instruments have been proposed to select women for bone densitometry, but no validated instruments are currently available for men. This study aims to address this deficiency by developing and validating a Male Osteoporosis Screening Tool (MOST) for Chinese men. Two thousand ambulatory men, aged 65 and above, were recruited from the general community in Hong Kong, and a cohort of 1,970 men with valid total hip and lumbar spine dual-energy X-ray absorptiometry (DXA) measurements was included in the current analysis. A 60% random sample was selected as the training sample for developing the screening tool, and the remaining 40% constituted the validation sample. Logistic regression and receiver operating characteristic (ROC) analysis were used to identify the simplest combination of risk factors to be included in the screening tool for predicting osteoporosis at the femoral neck, total hip, or lumbar spine. Body weight and quantitative ultrasound index (QUI) were found to contribute significantly to the area under the ROC curve (AUC), yielding an AUC of 0.823 in the training sample. The resulting MOST had a sensitivity of 94% and a specificity of 46% when using a cutoff score of 3. MOST had an AUC of 0.839 in the validation sample. The risk of osteoporosis was 1% among those with MOST scores < or = 2, but 72% among those with MOST scores > 7. Using a cutoff of 3, the negative predictive value was 97.5% which suggests that the 42% with MOST scores < or = 3 may be accurately screened out as being without osteoporosis, thus saving two fifths of our DXA resources. The positive predictive value was 72% when using a cutoff of 7, implying that MOST cannot replace DXA for case-finding purposes. Nevertheless, for resource allocation and patient satisfaction, it is prudent and economical to offer DXA screening first to the 6% with MOST scores > 7.