This paper examines changes in broom snakeweed populations (Gutierrezia sarothrae [Pursh] Britt. & Rusby) from 1979 to 2014 at three prairie grassland sites in New Mexico. Data gathered each fall were used to study broom snakeweed population dynamics and to estimate the probability that the relatively short-lived subshrub will die off or invade blue grama (Bouteloua gracilis [H.B.K. Lag]) rangelands. Annual broom snakeweed standing crop data were used to categorize populations as None (<100 kg ha^-1), Light (<300), Moderate (<750), or Heavy (≥750). Ordered logit regression was then used to estimate the frequency of transition between these categories over time depending on environmental and site factors. Significant variables found to influence annual variation in broom snakeweed included the broom snakeweed standing crop and density observed the previous period ( effect for continued broom snakeweed); grass standing crop the previous period (-); rainfall received from April to June ( ); and average temperatures during April ( ) and June (-). The probability of broom snakeweed invading an area that is currently without the plant ranges from about 1% to > 40% depending on environmental conditions and the amount of grass standing crop present. Transition probability estimates were also used in a Monte Carlo simulation model to evaluate the economics of broom snakeweed control. The economics of chemical broom snakeweed control were most strongly related to the rate of snakeweed reinvasion on treated areas and to the probability of natural die-off if infested areas were not sprayed.