[en] Recent in-pile slow TOP (≤1%P_o/s, P_o is nominal power) test results in CABRI and EBR-II were evaluated together with former results in TREAT and SLSF. These slow TOP database with various fuel design and preirradiation conditions gave a general understanding of failure mechanism and failure threshold dependency on fuel conditions. Fuel smear density is a key parameter for failure threshold and low to intermediate smear density provides high failure threshold. Fuel burnup and/or pre-irradiation linear power rating also play an important role for fuel swelling where its contribution for failure becomes considerable with high smear density fuel. Study on coolability of failed fuel pins was also performed with these failure data. Although very limited test data with continued overpower condition after failure made clear that a conclusion was difficult, the amount of ejected molten fuel into the coolant channel is judged to be the most important parameter for coolability. (author)