Creating　qualitative　measures　of　driver　workload　is　a　major　challenge　associated　with　matching　road,　transportation,　and　environmental　conditions　to　driver　ability.　Furthermore,　humanizing　road　safety　should　reduce　accidents.　This　study　had　two　parts:　seventeen　drivers　were　tested　in　a　simulator　cabin　that　emulated　a　scenic　which　is　the　same　as　a　test　highway.　Six　drivers　were　tested　on　a　real　highway.　Driver　workload　was　assessed　by　measuring　the　changes　in　heart　rate　and　pupil　size.　We　found　road　alignment　and　driver　workload　are　inversely　correlated　at　certain　speeds.　We　divided　driver　workload　into　three　levels:　comfortable,　moderate　tension,　and　high　tension.　Regression　analysis　was　used　to　determine　the　differences　among　levels.　Despite　similar　conditions,　a　difference　in　driver　workload　between　the　simulation　and　the　real　highway　was　observed.　The　results　were　used　to　obtain　three　levels　of　control　criteria　for　pupil　size　and　heart　rate,　which　was　used　to　evaluate　driver　workload.　©　2011　ASCE.