AbstractWe computed the impact solutions of the potentially dangerous Near Earth Asteroid (NEA) 2001 BB16 based on 47 optical observations from January 20.08316 UTC, 2001, through February 09.15740 UTC, 2016, and one radar observation from January 19.90347 UTC, 2016. We used two methods to sample the starting Line of Variation (LOV). First method, called thereafter LOV1, with the uniform sampling of the LOV parameter, out to LOV = 5 computing 3000 virtual asteroids (VAs) on both sides of the LOV, which gives 6001 VAs and propagated their orbits to JD2525000.5 TDT=February 12, 2201. We computed the non-gravitational parameterA2=(34.55±7.38)·10–14 au/d2 for nominal orbit of 2001 BB16 and possible impacts with the Earth until 2201. For potential impact in 2195 we find A2=20.0·10−14 au/d2. With a positive value of A2, 2001 BB16 can be prograde rotator. Moreover, we computed Lyapunov Time (LT) for 2001 BB16, which for all VAs, has a mean value of about 25 y. We showed that impact solutions, including the calculated probability of a possible collision of a 2001 BB16 asteroid with the Earth depends on how to calculate and take into account the appropriate gravitational model, including the number of perturbing massive asteroids. In some complicated cases, it may depend also on the number of clones calculated for a given sigma LOV1. The second method of computing the impact solutions, called thereafter LOV2, is based on a non-uniformly sampling of the LOV. We showed that different methods of sampling the LOV can give different impact solutions, but all computed dates of possible impacts of the asteroid 2001 BB16 with the Earth occur in accordance at the end of the 22nd century.