The paper presents the cancer mortality and incident ecological studies due to the natural background radiation for both women and men groups in all voivodeships in Poland in 2012. The potential correlation between local annual effective dose rates from natural sources and cancer mortality ratio (CMR), cancer incident ratio (CIR) and relative risk (RR) were analysed. All results are generally inconsistent and show both increase and decrease of CMRs, CIRs and RRs. However, the presented contradictions in trends of mortalities and incidents are statistically non-significant. Additionally, the data were adjusted to six confounding factors: people aged 70 or older, number of regular daily smokers etc. Based on these results, there is no regular trend observed for cancers related to natural background radiation in Poland, thus the popular linear no-threshold (LNT) hypothesis is not an accurate description of cancer risk prediction related to natural background radiation.