OBJECTIVE Sputum acid-fast bacilli (AFB) microscopy services are not available in all health facilities. Alternative procedures are needed to transport sputum samples to the diagnostic centres for detection of AFB. The objective of the present study was to evaluate sputum smears made by pot-method with the direct smears made immediately after sputum collection by Ziehl-Neelsen (ZN) method. METHODS Ninety three sputum samples from 49 pulmonary tuberculosis suspects were studied. Their direct smears (ZN smears) were stained by hot ZN method. The samples were then mixed with phenol ammonium sulphate basic fuchsin solution and stored at ambient conditions. The smears (pot smears), made on day 7, were then, decolourized and counter-stained for detection of AFB (pot method). The ZN and pot smears were read blind. After excluding 18 samples for various reasons, the results of pot and ZN smears of 63 samples from smear positive (2 of 3 direct smears were positive) and 12 from smear negative (3 of 3 direct smears were negative) patients were analysed. ZN method was the gold standard. RESULTS Pot and ZN smears were positive in 61 of 63 samples from smear-positive patients and negative in 11 of 12 smear-negative patients (kappa = 0.87). The sensitivity and specificity of pot method were 96.8 and 91.7 per cent respectively. CONCLUSIONS Sputum samples can be stored for up to seven days in the sputum container with phenol ammonium sulphate basic fuchsin solution. However, a comprehensive study needs to be done confirm the accuracy of the pot method for storage and transportation of sputum to microscopy centres for detection of AFB.