Trichinella surveillance data in Germany show for indoor housed pigs hardly any cases. Nevertheless, obligatory testing is in place for each slaughtered pig. According to EU legislation systematic Trichinella testing can be replaced by a risk-based surveillance system if the risk of Trichinella infection in fattening pigs is negligible. The probability to detect a positive herd (herd sensitivity) was taken as an indicator for the effectiveness of the surveillance. Four different diagnostic methods: a) digestion method, b) E/S-ELISA, c) Western Blot, and d) ELISA sequentially combined with Western Blot, were compared regarding herd sensitivity and specificity for different herd and sample sizes and different levels of prevalence. In a further step three potential surveillance systems were compared with regard to their suitability for herd classification: (i) classical Trichinella examination by artificial digestion method, (ii) ELISA screening followed by classical Trichinella examination and (iii) ELISA screening followed by Western Blot. Results show that: 1) testing by the artificial digestion method (i) provides only low sensitivity of detection for positive herds at present levels of prevalence despite perfect specificity. 2) The ELISA alone provides a high sensitivity of detection even at low sample sizes but at the cost of a very low herd specificity, converging towards zero at increasing sample sizes. In surveillance system (ii), a large number of farms would still need to be tested with the classical digestion method, as they would be misclassified as positive by the ELISA. 3) The Western Blot as well as ELISA screening followed by Western Blot offer a high probability of correct herd classification. The latter diagnostic system appears to be the most suitable for a risk based surveillance (iii) and provides--despite reduced sample sizes--a higher probability for a correct herd classification than the traditional Trichinella examination.