［目的］了解奶牛结核病在陕西省某县的流行情况。［方法］采用横断面研究方法，对该地区存栏量30头以上的规模奶牛场进行测定，估算其个体和群体阳性率，并绘制阳性场空间分布图。目标场是陕西省某县奶牛存栏量≥30头的奶牛场（共20个）。选择两阶段抽样策略：第一阶段，选择所有的20个奶牛场；第二阶段，计算选定奶牛场抽样量并随机选择抽样个体。抽样量采用估计流行率方法，通过Win Episcope 2.0软件计算，设定置信水平（CL）、预期流行率（P）和可接受误差（E）分别为95%、3%和2%。试验方法（PPD皮试）及阳性结果的判定标准参照《动物结核病诊断技术》（GB/T 18645—2002）执行。［结果］该县规模场奶牛结核病个体阳性率为1.7%，群体阳性率为60%。因检测方法特异性偏低，无法得到真实流行率。［结论］通过PPD皮试方法测得陕西省某县规模化奶牛场个体阳性率为1.7%，牛群阳性率为60%。阳性场空间分布相对聚集，均沿公路和河流分布。调查结果提示，在今后的牛结核病检测工作中，应充分考虑到诊断试验的特异性，并持续做好奶牛结核病的监测和净化工作。
[Objective]To determine the epidemic status of cow tuberculosis（TB）in a county of Shaanxi province. [Methods]A cross-sectional study for the dairy farms which had 30 or more cows was carried out to estimate the prevalence at the individual and herd level，and spatial distribution map of the positive farms was drawn. A total of 20 farms were targeted conforming to the above mentioned criteria and two-stage sampling strategy was adopted. Stage 1，selecting all the 20 dairy farms. Stage 2，calculating the sampling size and randomly sampling these experimental individuals from. Sample sizes were calculated by Win Episcope 2.0 by means of estimating prevalence，when parameters of confidence level（CL），expected prevalence（P）and acceptable error（E）were defined as 95%，3% and 2%，respectively. The test method（PPD skin test）and criteria of positive results were performed according to the Diagnostic Techniques for Tuberculosis of Animal（GB/T 18645—2002）. [Results] The estimated apparent prevalence of TB was 1.7% of individual level and 60% of herd level. However，the true prevalence could not be calculated in this study due to the low specificity of detection method. [Conclusion] By the method of PPD skin test， the individual positive rate of TB in that county was 1.7%，and it was 60% at herd level. Positive farms distributed rather?closely，usually along the main roads and rivers. The results indicated the diagnostic specificity should be taken into full account in the future. Meanwhile，long-term surveillance and eradication for bovine tuberculosis should be persisted in.