Leprosy control services face the problem of leprosy patients being misclassified by the lack of or the poor quality of skinsmear examination services. Misclassification increases the risk of relapse due to insufficient treatment if a multibacillary (MB) patient is classified as paucibacillary (PB), thereby also prolonging the time that the patient is infectious. The World Health Organization (WHO) recommends at present an alternative classification based on the number of skin lesions. Its reliability, however, has been questioned. Our investigation sought to determine the usefulness of the ML Dipstick, a simple field assay to detect IgM antibodies to phenolic glycolipid-I of Mycobacterium leprae, for the classification of leprosy patients in addition to lesion count. In this study, 264 leprosy patients were investigated. Of 130 patients with a positive bacterial index (BI), 19 (14.6%) had less than 6 lesions and would have been classified as PB. Out of 134 patients with a negative BI, 26 (19.4%) had 6 or more lesions and would have been classified as MB patients if the lesion counting system would apply. Thus, the classification based on the number of lesions only was found to be 85% sensitive and 81% specific (using the BI as the gold standard) at detecting MB cases among the studied population. Sensitivity would have increased if patients would have been classified according to a combination of the number of lesions and the dipstick result. In that case patients are classified as MB when they are either dipstick positive (N = 16), have more than 6 lesions (N = 43), or both (N = 94). Patients negative for both dipstick and number of lesions would have been classified as PB (N = 111). The classification based on the number of lesions alone left 19 BI-positive cases classified as PB, while the combination method of the ML Dipstick and number of lesions left only 8 BI-positive cases classified as PB (5 borderline, 2 borderline lepromatous and 1 tuberculoid), thus preventing undertreatment. The combination method of the ML Dipstick and lesion counting was found to be 94% sensitive and 77% specific, which is an improvement of 9% (chi-squared test, p = 0.025) in sensitivity compared to lesion counting only. The results of this study indicate that testing all patients initially classified by lesion counting as PB (48% in our study population) with the dipstick can significantly contribute to improved classification of leprosy patients for treatment purposes.