Empathy is a concept associated with various positive outcomes. However, to measure such a multifaceted concept, valid and reliable tools are needed. Negatively worded items (NWIs) are suspected to decrease some psychometric parameters of assessment instruments, which complicates the research of empathy. Therefore, the aim of this study was to assess the factor structure and validity of the TEQ on the Czech population, including the influence of the NWIs. Data were collected from three surveys. In total, 2239 Czech participants were included in our study. Along with socio-demographic information, we measured empathy, neuroticism, spirituality, self-esteem, compassion and social desirability. NWI in general yielded low communalities, factor loadings and decreased internal consistency. Therefore, in the next steps, we tested the model consisting of their positively reformulated versions. A higher empathy was found in females, married and religious individuals. We further found positive associations between empathy, compassion and spirituality. After the sample was split in half, exploratory factor analysis of the model with reformulated items was followed by confirmatory factor analysis (CFA), which supported a unidimensional solution with good internal consistency: Cronbach’s α = 0.85 and McDonald’s ω = 0.85. The CFA indicated an acceptable fit χ2 (14) = 83.630; p < 0.001; CFI = 0.997; TLI = 0.995; RMSEA = 0.070; SRMR = 0.037. The Czech version of the TEQ is a valid and reliable tool for the assessment of empathy. The use of NWIs in Czech or in a similar language environment seems to be questionable and their rewording may represent a more reliable approach.