Spatial-numerical associations (SNA) have been studied in various numerical and non-numerical tasks. Classically, participants respond faster with their right hand to larger numbers and faster with their left hand to smaller numbers even when number identity is irrelevant. SNAs are often explained as activation of a left-to-right mental number line. SNAs may reflect magnitude processes, relative order, or both. In the present research, I explored the SNA in an order judgment task. In three experiments, participants judged whether three digit sequences were ordered or unordered. People with more efficient calculation skill are faster and more accurate on order judgments than those with less efficient skills. Thus this task is assumed to capture important aspects of the mental representation of numerical sequences. I hypothesized that participants would show a SNA similar to that in other number tasks (e.g., parity, number comparison), that is, faster right hand responses to number sequences that are assumed to activate the right side (e.g., 6 7 8) and faster left hand responses to number sequences that are assumed to activate the left side (e.g., 2 3 4) of the mental number line. In three experiments, ordered but not unordered sequences showed consistent evidence for a typical SNA. In Experiment 1, Asian-educated individuals showed a stable SNA effect on both ascending and descending sequences whereas Arabic-speaking participants showed no significant SNA effects. English-speaking participants showed a typical SNA on descending sequences and a reversed SNA on ascending sequences. In Experiments 2 and 3, individuals with better arithmetic fluency showed a typical SNA. In Experiments 2 and 3, analyses using a multi-factor ANOVA design did not show evidence for SNARC effects. Instead, SNARC effects were found only using slope analyses on differences in responses with the right versus the left hand. The results are discussed in relation to the processes involved in sequence judgments and how those might differ from the processes involved in other numerical tasks. In summary, the SNA was elusive, even though participants were quite sensitive to number magnitude and the task itself required order judgments, which presumably should have activated the mental number line.